Feedback Loop Analysis

Feedback loops are circular chains of cause and effect where the output of a system influences its own input. Understanding them is essential for diagnosing persistent problems and designing effective interventions.

The Basics

What is a Feedback Loop?

A → B → C → A

The output (C) influences the input (A), which changes B, which changes C, which changes A again. The cycle continues.

Two Types of Loops

Reinforcing loops amplify change. They make things grow or decline exponentially.

• Success breeds success
• Debt creates more debt
• Panic causes more panic
• Growth enables more growth

Balancing loops resist change. They stabilize systems around a target.

• Thermostat maintains temperature
• Hunger prompts eating
• Inventory triggers reordering
• Fatigue forces rest

Why Loops Matter for Diagnosis

Persistent Problems

If a problem keeps coming back despite interventions, look for a reinforcing loop that's working against you.

Example: High turnover

• Turnover increases workload on remaining staff
• Increased workload leads to burnout
• Burnout leads to more turnover

Breaking this loop requires intervention at multiple points, not just hiring faster.

Unexpected Resistance

If a change doesn't stick, look for a balancing loop that's resisting it.

Example: Process improvement

• New process is introduced
• Extra effort required to follow new process
• Under deadline pressure, people revert to old habits
• System returns to previous state

The balancing loop around "minimize effort under pressure" defeats the change.

Delayed Effects

Loops often have delays. Today's cause produces tomorrow's effect, making the connection hard to see.

Example: Quality investment

• Invest in quality → improved products (6-month delay)
• Improved products → higher customer satisfaction (3-month delay)
• Higher satisfaction → more referrals (6-month delay)
• More referrals → higher revenue (3-month delay)
• Higher revenue → more investment capacity

Total loop time: 18 months. In the meantime, the investment looks like pure cost.

Analysis Process

Step 1: Map the Variables

Identify all the factors involved in the system. Don't worry about connections yet.

Step 2: Trace the Connections

For each variable, ask: "What does this directly influence?"

Draw arrows from cause to effect. Label each arrow:

• + if an increase in the cause leads to an increase in the effect
• - if an increase in the cause leads to a decrease in the effect

Step 3: Identify Loops

Follow the arrows around. When you return to a starting point, you've found a loop.

Determine loop type:

• Count the negative (-) arrows in the loop
• Even number (including zero) = Reinforcing loop
• Odd number = Balancing loop

Step 4: Find Delays

For each connection, estimate the delay. Significant delays change how the loop behaves and how interventions should be designed.

Step 5: Identify Leverage Points

Where in the loop can you intervene?

High leverage points:

• Where you can break a reinforcing loop that's causing harm
• Where you can strengthen a balancing loop that's maintaining dysfunction
• Where you can change the direction of an arrow

Intervention Strategies

Breaking Reinforcing Loops

You don't need to break every link. Find the weakest point and intervene there.

Example: For the turnover loop, possible interventions:

• Reduce workload impact (cross-training)
• Increase burnout resilience (support programs)
• Speed up replacement (hiring pipeline)
• Change the base state (proactive staffing)

Modifying Balancing Loops

To enable change, you need to shift what the loop is balancing toward.

Example: For the process adoption loop, possible interventions:

• Change the target (make new process the expected default)
• Reduce resistance (make new process easier)
• Add accountability (create cost for reverting)
• Allow the transition (temporary performance dip expected)

Adding New Loops

Sometimes the solution is to create a new loop that counteracts the problematic one.

Common Patterns

The Fixes That Fail

• Problem occurs
• Quick fix applied
• Symptoms subside
• Root cause unaddressed
• Problem returns, often worse

The quick fix creates a balancing loop that masks the problem without solving it.

Shifting the Burden

• Problem occurs
• Symptomatic solution applied
• Side effect undermines fundamental solution
• Dependency on symptomatic solution grows
• Capability for fundamental solution atrophies

Example: Outsourcing expertise instead of building it internally.

Eroding Goals

• Performance falls short of goal
• Pressure to close the gap
• Goal is lowered instead of performance raised
• New goal becomes the standard
• Repeat

The balancing loop works, but toward the wrong target.

Systems resist change. Understanding why tells you where to push.